x
BACK-AND-FORTH HERMAPHRODITISM:
PHYLOGENETIC CONTEXT OF REPRODUCTIVE
SYSTEM EVOLUTION IN SUBDIOECIOUS
DAPHNE LAUREOLA
Conchita Alonso1,2 and Carlos M. Herrera1
1 Estació n
Biológica de Doñana, Consejo Superior de Investigaciones Cientı́ficas (CSIC). Apdo 1056, 41080, Sevilla, Spain
2 E-mail:
conalo@ebd.csic.es
Recent phylogenetic analyses of sexual reproductive systems supported the evolutionary pathway from hermaphroditism to
dioecy via gynodioecy in different groups of angiosperms. In this study, we explore the evolution of sexual reproductive systems
in Daphne laureola L. (Thymelaeaceae), a species with variation in reproductive system among population. Sequences from the
ITS region of the nuclear ribosomal cistron and two plastid markers (psbA-trnH and ndhF) were analyzed and used to map the
population reproductive system along the molecular phylogeny. Our results support D. laureola as a monophyletic lineage with
three different clades within the Iberian Peninsula. The hermaphroditic populations belong to two different clades, whereas
gynodioecy is ubiquitous but characteristic of the third clade, which grouped together all the North-Western Iberian populations
sampled, including the apparently oldest haplotype sampled. Gynodioecy appears as the most likely basal condition of the 13
analyzed populations, but different evolutionary transitions in reproductive sexual system were traced within each D. laureola
clade. Both ecological conditions and (meta)population dynamics may help explain plant reproductive system evolution at the
microevolutionary scale. Phylogenetic studies in which the historical relationships between populations differing in reproductive
system can be ascertained will help to clarify the process.
KEY W ORDS:
Geographic variation, gynodioecy, Iberian Peninsula, ITS, microevolution, molecular markers, phylogeography.
Hermaphroditism is considered the ancestral condition from
which all other sexual reproductive systems have evolved in
the angiosperms (see Barrett 2002 for a review). Gynodioecy,
where hermaphrodite and female individuals coexist within populations, has been traditionally viewed as one intermediate evolutionary stage between the two most common reproductive systems, hermaphroditism and dioecy. The evolutionary pathway
from hermaphroditism to dioecy via gynodioecy has been satisfactorily modeled theoretically, and requires the spreading of
male-sterile mutants into cosexual populations followed by selection against female function in the remaining hermaphrodites
(Schultz 1994; Charlesworth 1999). At the first stage, to spread
and persist within populations, the male-sterile mutants must have
some consistent fecundity advantage that compensates for their
gametic disadvantage. In theory, the female frequency within single populations of infinite size would be regulated by the relative
seed production of each sexual morph, and the degree of selfing
and inbreeding depression (Charlesworth 1999). Metapopulation
models have further indicated that whenever a population is subdivided into local groups, individual fitness will be also a function
of local morph frequency and pollen limitation of seed production, which would determine if evolutionary processes allow the
maintenance of polymorphism (McCauley and Taylor 1997). The
stability of gynodioecy can be also favored by nucleo-cytoplasmic
sex determination (Bailey et al. 2003; Bailey and Delph 2007;
Dufay et al. 2007; but see Schultz 1994).
In the last two decades important efforts have been made
to verify these models, explain some seeming exceptions to the
rule, and understand the ecological conditions favoring such microevolutionary transitions in sexual reproductive system (Delph
and Carroll 2001; Medrano et al. 2005; Ashman 2006; McCauley
and Bailey 2009). In addition, some phylogenetic analyses of
reproductive system evolution based on molecular markers (see
Weller and Sakai 1999 for a review; Weiblen et al. 2000; Vamosi
et al. 2003; Navajas-Pérez et al. 2005) have also supported the
theoretical pathway for a few plant groups in a macroevolutionary context, and reversals from dimorphism to hermaphroditism
have been also inferred at the interspecific level (e.g., Weller et al.
1995; Desfeux et al. 1996; Weiblen et al. 2000).
Links between the micro and macroevolutionary approaches
are being provided by studies of geographical variation in species
that exhibit among-population variation in reproductive system
(hereafter referred to as subdioecious). These investigations have
shown that environmental stress, founder effects, and interactions
with pollinators, are commonly correlated with the frequency of
females in subdioecious species (e.g., Case and Barrett 2004;
Nilsson and Ågren 2006; Ramsey et al. 2006; Alonso et al. 2007;
Caruso and Case 2007). Finally, intraspecific phylogeographic
studies in which the historical relationships between populations
differing in reproductive system can be ascertained using molecular markers may also provide crucial information to identify
the most common transitions in reproductive system evolution
at the intraspecific level. By mapping selected characters on a
phylogeny, such studies can suggest whether other morphological, functional, and ecological characters have played a role in
this microevolutionary process. However, such studies are scarce
(see e.g., Fénart et al. 2006), most likely because of the difficulties in finding well-supported intraspecific phylogenetic trees and
intraspecific variation in reproductive system simultaneously for
the same species.
Our study uses a phylogenetic approach to provide
a historical frame for the evolutionary transitions between
hermaphroditism and gynodioecy at an intraspecific level. Daphne
laureola L. (Thymelaeaceae) is a shrub widely distributed
in the understory of European forests. Monomorphic (i.e.,
hermaphroditic) and gynodioecious populations are patchily distributed over the Iberian Peninsula (Alonso et al. 2007). Gynodioecious populations prevail in the northwest and southeast of
this large geographic area and, most likely, also at other disjunct
areas of the species range (gynodioecious populations occur also
in Slovenia and Morocco; C. Alonso, unpubl. data). In addition,
there are at least two disjoint, distant areas in the northeast and
southwest of the Iberian Peninsula in which pure hermaphroditic
populations exist. We attempt to understand such a geographic
pattern using molecular phylogenetic tools. Combining different
marker systems is usually advantageous to unravel intraspecific
evolutionary histories; thus we analyze the internal transcribed
spacer (ITS) region of the nuclear ribosomal cistron, and two
plastid markers (psbA-trnH and ndhF). Nuclear and plastid markers show contrasting modes of inheritance. It is thus particularly
interesting to combine both to track plant reproductive system
evolution because, as mentioned above, gynodioecy is frequently
determined by the interaction between maternally inherited cytoplasmic male sterility (CMS) factors and biparentally inherited nuclear genes that restore male fertility (Kaul 1988; but see
McCauley and Olson 2008).
The following questions are addressed in this article. What
are the phylogenetic relationships between D. laureola populations from different Iberian regions and how does the sexual reproductive system change along the molecular phylogeny? Are
D. laureola populations from the hermaphroditic disjunct areas
genetically more related to each other rather than to gynodioecious populations? If so, is hermaphroditism the ancestral trait
among D. laureola populations? Or alternatively, in which direction and how many times has the transition between gynodioecy
and hermaphroditism occurred?
Materials and Methods
PLANT MATERIAL AND STUDY SITES
Daphne laureola L. (Thymelaeaceae) is an early-season flowering, insect-pollinated and bird-dispersed, evergreen shrub widely
distributed in Europe (Brickell and Mathew 1976). Typically it
grows in the undergrowth of shady mountain forests. In the Iberian
Peninsula, the species is frequent in the northern Cantabrian and
Pyrenean Mountains, characterized by Atlantic climate and deciduous and mixed forest, and also in the southern Mediterranean
Betic Ranges, where it inhabits evergreen sclerophyllous and subsclerophyllous woodlands (Alonso et al. 2007). At this large geographic scale, D. laureola exhibits among-population variation
in reproductive system. Gynodioecy is the most common sexual
system and prevails in the northwestern and southeastern Iberian
populations but there are at least two disjunct, distant areas in the
northeast and southwest of the Iberian Peninsula in which purely
hermaphroditic populations exist (Alonso et al. 2007).
We studied the phylogenetic relationship between individuals from four hermaphroditic and seven gynodioecious Iberian
populations of D. laureola (see Fig. 1 for locations). Those populations were a subsample of a broader survey conducted to determine the geographic variation in the proportion of female and
hermaphrodite plants per population (Alonso et al. 2007). All
of them have more than 100 individuals. Gynodioecious populations were located in NW, NE, and S Iberian regions and their
frequency of females (11%–54%; Table 1) encompassed the range
Daphne laureola L. distribution map according to Meusel et al. (1978). Locations of the two Slovenian populations are marked
in the distribution map and the 11 Iberian populations sampled are shown in more detail, with the three a priori geographical regions
we distinguished indicated. Black and white dots denote gynodioecious and hermaphroditic populations, respectively, population codes
F ig u r e 1.
as in Table 1.
of variation recorded at the broad geographic scale (see Alonso
et al. 2007 for details). Three hermaphroditic populations were
selected from the NE region where they are frequent, and one
from the S region where they are not so common. In addition,
we included two European populations outside the Iberian Peninsula, one hermaphroditic and one gynodioecious, to better char-
acterizing the phylogeographic relationships by broadening the
geographic distances (Fig. 1). The two populations were located
in two different Slovenian mountain ranges, distant about 70 km
from each other, and the hermaphroditic population was smaller
(22 individuals) than the gynodioecious population (>100 individuals).
Location, reproductive system, female percentage, ribotypes (ITS) and haplotypes (psbA-trnH and ndhF sequences combined)
of the Daphne laureola populations studied. Regions and population codes as in Figure 1.
Ta b l e 1 .
Region
Population (code)
Latitude N/Longitude W
Reproductive system
% females
Ribotype
Haplotype
S Iberia:
Huerta Vieja (HV)
Cuevas Bermejas (CB)
La Maroma (ST)
Hoyos de D. Pedro (CA)
Las Cruces (GA)
Bosque Peloño (BP)
Aldatz (AL)
Selva Villanúa (SV)
S. Juan de la Peña (SJ)
Gresolet (GR)
Tagamanent (TG)
ca. Vrhnika (SL1)
ca. Kocevje (SL3)
38.12/−2.81
37.96/−2.85
36.91/−4.02
36.10/−5.51
42.59/−7.04
43.17/−5.13
43.02/−1.86
42.68/−0.50
42.51/−0.67
42.26/1.72
41.75/2.30
45.96/14.19
45.67/14.99
Gynodioecy
Gynodioecy
Gynodioecy
Hermaphroditism
Gynodioecy
Gynodioecy
Gynodioecy
Gynodioecy
Hermaphroditism
Hermaphroditism
Hermaphroditism
Hermaphroditism
Gynodioecy
35.5
32.2
10.9
0
39.4
36.0
54.0
54.0
0
1.0
0
0
15.0
A
A
C
C
D
D
D
D
A
A
A
B
A/B
3/4
3
3
5
1
1
1
1/2
6
6
6
6
6
NW Iberia:
NE Iberia:
Slovenia:
Each population was represented in the phylogenetic analyses by two individuals of each sex (N = 42 individuals) to allow
detecting possible polymorphisms linked to sex (e.g., McCauley
and Olson 2003). The small sample size per population was chosen on the assumption of reduced levels of polymorphism in DNA
sequences at the within-population level, as frequently found for
plastid markers (e.g., Cornman and Arnold 2007).
Finally, leaf samples from two to six individuals of other
six European Daphne species (D. blagayana, D. cneorum, D.
gnidium, D. mezereum, D. oleoides, and D. rodriguezii) were
also collected and similarly analyzed to allow the rooting of the
phylogenetic trees because no molecular phylogeny of Daphne is
currently available.
DNA ISOLATION AND MOLECULAR ANALYSIS
Total genomic DNA was isolated from freshly frozen bud tissue
maintained at −80◦ C, using a DNeasy 96 Plant Kit (QIAGEN,
Inc., Valencia, CA). We selected one nuclear and two cpDNA regions for sequencing: the ITS region of the 18S-5.8S-26S nuclear
ribosomal cistron and the psbA-trnH intergenic spacer and the
ndhF gene, respectively.
PCR-amplification was performed by using PCR-Beads kits
(PuReTaq Ready-To-GoTM , Amersham Biosciences, UK) with
approx. 25 ng of DNA template, and 0.2 mM of each primer
in a final volume of 25 μL. The following primers were used
for each region: P1A/P4 for the complete ITS region (ITS1,
5.8S gene, ITS2; White et al. 1990; Downie and Katz-Downie
1996), psbA/trnH for their intergenic spacer (Sang et al. 1997),
and primers 1318 and 2110R for the 3i end of the ndhF plastid
gene (Olmstead and Sweere 1994). The PCR mix underwent the
following conditions on a 9700 thermal cycler (Perkin–Elmer,
Norwalk, CT): 10 min denaturing at 95◦ C, 37 cycles of 30 s denaturing at 95◦ C, 30 s annealing at 52◦ C, and 1 min extension at
72◦ C and a final extension step at 72◦ C for 7 min. The PCR products were then purified using spin filter columns (UltraCleanTM
PCR Clean-upTM Kit, MoBio Laboratories, Inc., Carlsbad, CA)
following the protocols provided by the manufacturer and directly
two-way sequenced with an ABI Prism BigDye Terminator Cycle
kit on a ABI Prism 3730 (Applied Biosystems, Norwalk, CT).
The resulting electropherographs were manually proofread and
the complementary strands combined into one sequence to identify ambiguous positions assisted by Sequencher 4.8 (Gene Codes
Corporation 2007).
DATA ANALYSES
Sequences and alignment
Forty-two ITS sequences of D. laureola and 18 of the other
Daphne species analyzed were generated from forward and
reverse sequences (GenBank accession numbers GQ167491–
GQ167548, HQ268822–HQ268823). Forty-two sequences of the
psbA-trnH marker were obtained for D. laureola, and 16 of the
other Daphne species (GenBank accession numbers GQ167433–
GQ167490). Thirty-seven sequences of the ndhF marker were
obtained for D. laureola, and 10 of the other Daphne species (GenBank accession numbers GQ167388–GQ167432, HQ268820–
HQ268821). For sequence analyses, multiple alignments were
performed using ClustalX, with default parameters for gap opening and extension (Thompson et al. 1997). The ITS and ndhF
regions analyzed did not show any indel variation within the D.
laureola sequences that constitute the focal purpose of this study
(see Appendices S1 and S2). Consequently, gaps were treated as
missing data. The alignment of the psbA-trnH intergenic spacer
was not so straightforward. First, at the interspecific level a 5
bp inversion occurred at position 36, that was coded as a single character change, with all ingroup taxa exhibiting the same
character state for the inversion. Second, the initial alignment of
our study sequences (see Appendix S3) suggested that two insertion/deletion events occurred within the D. laureola sequences
analyzed that would be ignored if treated as missing data. The
first one involved an A rich region with 9, 10, or 11 repeats, and
the second one involved a TA motif with 7, 10, or 13 repeats.
The complex architecture of this plastid marker that can affect automatic alignment (Storchova and Olson 2007; Morrison 2009),
and the absence of sequences of other Thymelaeaceae phylogenetically close to Daphne that could improve the phylogenetic
interpretation of these events precluded us from coding indels
and thus gaps were conservatively treated as missing data.
Haplotype detection
Statistical parsimony network implemented in the TCS program
(Clement et al. 2000) was used to detect haplotypes, estimate the
mutational differences among them justified by the parsimony
criterion, and establish the resulting network that would allow
incorporating the frequently nonbifurcating genealogical information associated with population-level divergences. Confidence
level was set at 95%, and gaps were treated as missing data.
Phylogenetic analysis
Five ITS sequences of the sister genus Thymelaea (GaliciaHerbada 2006) retrieved from GenBank (AJ549483 T. villosa, AJ549489 T. argentata, AJ549468 T. dioica, AJ549470 T.
granatensis, and AJ549442 T. passerina), were included into the
analyses to explore the phylogenetic relationships of Daphne with
respect to another Thymelaeaceae and improve the rooting of our
focal sequences. Phylogenetic relationships among taxa were estimated using maximum parsimony analysis (unordered characters, equal weights; referred as MP hereafter) as implemented in
PAUP 4.0b10 (Swofford 2002). Heuristic searches with random
trees used as a starting point and 10 random addition sequences
of taxa were performed with the TBR branch-swapping algorithm
and MulTrees options in effect. Bootstrap support (referred as BS
hereafter) values were calculated on 1000 replicates.
Bayesian inference phylogenetic analyses were also conducted using MrBayes version 3.1.2 (Huelsenbeck and Ronquist
2001). Analyses were run for 100,000 generations with four
MCMC chains and two independent runs with trees sampled every 100th generation, and the first 80–90 sampled trees discarded
by burn-in. Modeltest version 3.7 (Posada and Crandall 1998) was
used to determine the DNA substitution model that best fitted the
data among those available in MrBayes. The model chosen for our
data using hierarchical likelihood ratio tests was HKY + G, which
assumes a time-reversible process, a nonuniform distribution of
nucleotides and different rates for transitions and transversions,
and accounts for among-site rate variation using the gamma distribution. Thus, we set Nst = 6 and rates = gamma options in
MrBayes analyses.
We performed maximum likelihood ancestral state reconstructions of the population reproductive system on the MP tree
based on the three sequenced regions analyzed simultaneously,
that was topologically similar but with a better resolution than
those obtained from cpDNA and rDNA markers analyzed independently. As reproductive system is a population-level feature,
we reduced our sample to just a single individual per population. This procedure should not affect phylogenetic signal in any
important way, because in most cases all the samples from the
same population were invariant in sequence and exhibited identical ribotypes and haplotypes (Table 1). In the only three populations where polymorphism occurred at some marker, variability
involved a single individual out of the four sampled, and we represented each population using the commonest variant. We used
Mesquite 2 (Maddison and Maddison 2007) and coded population
sexual system as a binary character (hermaphroditism vs. gynodioecy). The rate of a character’s evolution was estimated under the
simple Markov k-state one-parameter stochastic model, assuming
that transition between hermaphroditism and gynodioecy had the
same probability in both directions.
Results
DAPHNE LAUREOLA IN THE CONTEXT OF OTHER
EUROPEAN DAPHNE SP.
ITS-based phylogenetic analyses
The percentage of potentially parsimony-informative sites in the
complete ITS region analyzed was 19.2% from 604 characters.
The score of the best trees found was 246, and the strict consensus
tree had consistency index (C.I.) = 0.825 (Fig. 2). Both parsimony
and Bayesian analyses yield a similar topology in their resulting
trees where all D. laureola sequences strongly group together
in a single lineage (BS and posterior probability 100%). The
lineage splits into two separate lineages both with a high BS. One
encompasses the representatives from NW Iberian populations
including the two Pyrenean gynodioecious populations (AL and
SV), and the second groups the remaining sequences. This second
group remains largely unresolved due to lack of variation, except
for two small groups of sequences. The first group singles out (BS
86%) those sequences coming from the two westernmost Southern
Iberian populations (ST and CA), and the second groups with a
low BS (68%) several Slovenian sequences.
The results of TCS analysis identify four different ribotypes
A, B, C, and D in D. laureola (Table 1), A being the most frequent within the sequences analyzed and the one identified as the
oldest in the sample. Different Daphne species yielded different
unconnected networks (results not shown). One important feature is that all the samples from the same population invariantly
share the same ribotype, except for the gynodioecious Slovenian
population (Table 1). In this population three individuals have
ribotype A and the other one has ribotype B, an infrequent ribotype present only in the second Slovenian population studied
and differing by only a single change with regard to ribotype A.
Hermaphroditic populations in NE Iberian Peninsula all have ribotype A, which was also shared with the two easternmost Southern
Iberian populations (CB and HV) and three of four individuals
of the gynodioecious Slovenian site mentioned above. Ribotype
C differed from A in two changes and was only present in the
two westernmost Southern populations (ST and CA). Ribotype D
differed in nine changes with regard to ribotype A and was present
in the four gynodioecious populations studied in Northern Iberian
Peninsula.
Plastid DNA-based phylogenetic analyses
The percentage of potentially parsimony-informative sites was
4.1% from 758 characters in the ndhF region and 8.8% from 318
characters in the psbA-trnH intergenic spacer (after coding one
inversion of five characters as a single change). The two markers
suggested a similar topology and thus we combined their analyses,
discarding the individuals that lacked one of the two markers. The
score of the best trees found was 75 and the strict consensus tree
had C.I. = 0.947.
Plastid DNA based phylogenies also revealed D. laureola as
a monophyletic lineage where no haplotypes are shared between
different species (Fig. 3). The monophyly of the lineage is supported by a BS value and posterior probability of 100% (Fig. 3).
The study sequences appear included in a large polytomy where
all the NW Iberian individuals, singled out according to the ITS
analyses, are nested. Within this polytomy three lineages arise,
all supported by strong Bayesian posterior probabilities (Fig. 3):
the first one gathers NE Iberian and Slovenian populations (BS
99%), the second one includes most sequences from the Southern
Iberian populations, and the last one joins two single sequences
from two Southern Iberian populations (BS 85%).
DL23_ST
DL24_ST
DL25_ST
DL26_ST
DL35_CA
DL36_CA
DL38_SL3
DL41_SL1
DL42_SL1
DL01_TG
DL02_TG
DL03_GR
DL04_GR
DL05_SJ
DL06_SJ
DL27_CB
DL28_CB
DL29_CB
DL30_CB
DL31_HV
DL32_HV
DL33_HV
DL34_HV
DL37_SL3
DL39_SL3
DL40_SL3
DL07_SV
DL08_SV
DL09_SV
DL10_SV
DL11_AL
DL12_AL
DL13_AL
DL14_AL
DL15_GA
DL16_GA
DL17_GA
DL18_GA
DL19_BP
DL20_BP
DL21_BP
DL22_BP
D. gnidium
D. gnidium
D. gnidium
D. gnidium
D. gnidium
D. gnidium
D. cneorum
D. cneorum
D. cneorum
D. cneorum
D. mezereum
D. oleoides
D. oleoides
D. rodriguezii
D. rodriguezii
D. rodriguezii
D. blagayana
D. blagayana
T. dioica
T. granatensis
T. argentata
T. passerina
T. villosa
10.0
Phylogenetic relationships based on ITS sequence of the 42 D. laureola individuals from 13 populations (in black, population
codes as in Table 1) and the other six Daphne species analyzed in this study, together with five Thymelaea species retrieved from GenBank. Maximum Parsimony strict consensus tree (submitted to TreeBase <http://purl.org/phylo/treebase/phylows/study/TB2:S10851>)
obtained with branch lengths proportional to inferred DNA changes. Bootstrap values (>70) and Bayes posterior probabilities (>80)
noted near branches left and right of a slash, respectively. The scale of substitution per site is indicated.
F igu r e 2.
. oleoides
. cneorum
. cneorum
. cneorum
. gnidium
Phylogenetic relationships based on the two plastid regions studied (psbA-trnH intergenic spacer and ndhF gen) of the 37 D.
laureola individuals from 13 populations (in black, population codes as in Table 1) and other four Daphne species analyzed. Maximum
Parsimony strict consensus tree (submitted to TreeBase <http://purl.org/phylo/treebase/phylows/study/TB2:S10851>) obtained with
branch lengths proportional to inferred DNA changes. Bootstrap values (>70) and Bayes posterior probabilities (>80) noted near branches
left and right of a slash, respectively. The scale of substitution per site is indicated.
F ig u r e 3.
D. blagayana
D. cneorum
D. cneorum
D. gnidium
D. gnidium
D. gnidium
DL01 TG
DL03 GR
DL05 SJ
DL38 SL3
DL41 SL1
DL23 ST
Haplotype network of the study sequences based on
the two plastid regions studied (psbA-trnH intergenic spacer and
ndhF gen). The maximum number of steps connecting parsimoniously two haplotypes is indicated. The haplotype with the highest outgroup probability, that correlates with haplotype age, is
displayed as a square, whereas other haplotypes are displayed
as ovals whose size corresponds to the haplotype frequency in
the sample. A map is provided with the geographic distribution
of main D. laureola haplotypes within the Iberian Peninsula, the
Slovenian populations not shown in the map have haplotype 6.
Haplotypes only represented by a single individual each and the
populations bearing those are marked with an asterisk.
F igu r e 4 .
DL35 CA
DL28 CB
DL32 HV
DL08 SV
DL11 AL
DL16 GA
DL20 BP
D. mezereum
D. oleoides
Six haplotypes can be inferred from the TCS analysis of
D. laureola combined plastid sequences psbA-trnH and ndhF
(Table 1). None of them is present in any of the other sampled
species, for which the haplotype networks remain unconnected to
the main ingroup network (results not shown). Three haplotypes
(2, 4, and 5) are represented by only a single individual each and
thus should be interpreted with caution. Haplotype 1, the most
frequent one and plausibly so the oldest in our sample (Clement
et al. 2000), is characteristic of NW Iberian populations (Fig. 4),
except for a single individual in SV population with haplotype 2,
differing in a single change in the ndhF sequence. The Haplotype
3 characteristic of most of the Southern Iberian populations differs
from the previous one in one change in the psbA-trnH region. Haplotype 6 is characteristic of hermaphroditic NE Iberian and the two
Slovenian populations and differs from the previous ones in five
and six changes, respectively. Finally, Haplotype 5, present in the
CA Southern Iberian population is the most dissimilar one (Fig. 4).
MAPPING REPRODUCTIVE SYSTEM IN THE
PHYLOGENY
The MP tree based on a single individual per population and
the three sequenced regions analyzed simultaneously showed a
topology similar to those obtained using all the study samples,
D. rodrigueziii
D. rodrigueziii
Maximum likelihood ancestral-state reconstruction of
reproductive system evolution on the strict MP consensus tree obtained using the three sequenced regions simultaneously and a
single individual per population. White and black circles in the
tips denote hermaphroditic and gynodioecious populations, respectively. In the reconstructed internal nodes white and black display the proportional likelihood support for hermaphroditic and
gynodioecious states, respectively. Daphne laureola lineage and
the three different clades NW, NE (including the two Slovenian
populations), and S obtained are indicated. Population codes as in
Table 1.
F ig u r e 5.
with monophyly of D. laureola supported by 100% BS, and the
three major lineages distinguished within this clade (NW, S, and
NE; Fig. 5) being consistent with those obtained with all the
available samples (Figs. 2 and 3). None of the rest of species
showed evidence of gynodioecy (Brickell and Mathew 1976),
and consequently the most likely ancestor between D. laureola
and the closest species in the phylogeny was hermaphroditism
with 86% support (Fig. 5).
The hypothetical ancestor common to all the D. laureola
study populations showed an 81% likelihood of being gynodioecious (Fig. 5). Within D. laureola, the NW clade (91% BS) that
grouped all the Northern gynodioecious populations had a 100%
likelihood of having a common ancestor gynodioecious (Fig. 5).
The common ancestor to all S and NE populations had 89%
likelihood of being gynodioecious, hermaphroditism being thus
a derived condition in the CA Iberian population. Furthermore,
the NE clade (100% BS) had 100% likelihood of having a common ancestor hermaphroditic, gynodioecy being thus a derived
condition in the Slovenian SL3 population (Fig. 5).
Summing up, gynodioecy appeared as the most likely ancestral state common to all the D. laureola populations studied,
and the different evolutionary transitions in reproductive system
traced within each clade account for current geographic variation
at the within-species level (Fig. 5).
Discussion
PHYLOGEOGRAPHY OF D. LAUREOLA IN THE IBERIAN
PENINSULA
The Iberian Peninsula is a relatively isolated, physiographically
complex area in the western part of the Mediterranean Basin,
with strong climatic gradients between the Atlantic, Mediterranean and continental territories, and contrasted soil types. The
Iberian mountain ranges, that predominantly run east-west, have
apparently provided multiple Pleistocene glacial refugia for many
temperate European tree species, which currently retain a higher
genetic diversity than in more northern areas of their distribution
ranges (Gó mez and Lunt 2007; de Heredia et al. 2007). Such longterm historical processes might have affected even more strongly
the genetic diversity and phylogeography of an understory species
like D. laureola. The disjunct distribution of D. laureola within
the Iberian Peninsula (Fig. 1) coupled with the recent discovery of among-population variation in reproductive system in both
the Northern and Southern areas of the Iberian distribution range
(Alonso et al. 2007) prompted this phylogeographic study aimed
to describe the genetic relatedness between several populations
broadly distributed geographically and to unravel the evolutionary transitions between hermaphroditism and gynodioecy at the
intraspecific level.
Our results supported D. laureola as a monophyletic lineage.
No haplotypes or ribotypes were shared with different species.
At intraspecific level, in one hand we found out that the a priori
distinction between NE and NW regions, based on the reduced
abundance of D. laureola in the Eastern part of the Cantabrian
mountains, did not accurately reflect the phylogenetic history of
the species because the two gynodioecious populations sampled
in the Western Pyrenees (AL and SV) were genetically akin to
a well-supported NW clade based on ITS, characterized by ri-
botype D, and they also share the distinctive NW haplotype 1,
plausibly the oldest in our sample, with the only exception of a
single sequence with the infrequent haplotype 2 (Fig. 4). Furthermore, the plastid markers also indicated that all samples from the
hermaphroditic NE Iberian populations, including the one located
at a southern pre-Pyrenean foothill (SJ), were genetically interrelated and grouped together with the Slovenian samples, characterized by the distinctive haplotype 6 (Fig. 5). Thus, genetic
differentiation of D. laureola in the North apparently occurred at
the central region of the Pyrenees, somewhere between the prePyrenean (SJ) and Pyrenean (SV) populations, distant only a few
kilometers (Fig. 1), but with a sharp contrast in geological history and climate (Garcı́a-Castellanos et al. 2003). Interestingly,
genetic differentiation occurred together with disparity in population reproductive system, with the NE Iberian populations, characterized by drier and warmer locations, being all hermaphroditic
(Alonso et al. 2007) and more genetically related to Slovenian
populations than to the other Iberian populations sampled. On
the other hand, the Southern populations showed characteristic haplotypes 3, 4, and 5 and two different ribotypes, one of
them shared with the NE clade, suggesting also certain genetic
differentiation within this region, further supported by analyses
of amplified fragment length polymorphisms (AFLP) variation
(A. R. Castilla, C. Alonso and C. M. Herrera, unpubl. data). To ascertain whether genetic differentiation is due to different D. laureola refugia during the last glaciation in the Iberian Peninsula, like
in several of the forest tree species such as Fagus sylvatica (Magri
et al. 2006), deciduous oaks (Olalde et al. 2002), and evergreen
oaks (de Heredia et al. 2007), or to local adaptation and diversification would require a more comprehensive phylogeographic
sampling of the species. The remarkable diversity of ribotypes and
haplotypes already obtained with our limited geographic sample
apparently supports the hypothesis of multiple Iberian refugia.
However, it could be further related to variation in reproductive
system itself if gynodioecious species harbor more old haplotypes,
as recently found with mitochondrial genes (Touzet and Delph
2009).
EVOLUTION OF REPRODUCTIVE SYSTEM IN
D. LAUREOLA
A recent review of gender variation within Thymelaeaceae has
shown the existence of unisexual flowers and diverse reproductive systems in a large proportion of genera (Beaumont et al.
2006). Such variation was concentrated in the largest subfamily
Thymelaeoideae, to which most genera including Daphne belong, in contrast to the other three smaller subfamilies in which
hermaphroditism is ubiquitous. In particular, the sister genus
Thymelaea (Van der Bank et al. 2002; Galicia-Herbada 2006)
is highly variable in reproductive system, and in the Iberian
Peninsula only T. passerina of 21 species has a monomorphic
hermaphroditic reproductive system (Pedrol 1997). Although the
genus Daphne (ca. 70 spp) apparently is not so variable in reproductive system, gynodioecy is known from D. kamtchatica
(Kikuzawa 1989), and unisexual flowers have been occasionally
reported from D. acutiloba, D. jezoensis, D. mezereum, D. odora,
and D. oleoides (Brickell and Mathew 1976; Nieto Feliner 1997),
which suggests that “further study of the genus is required to determine the significance of these characters” (Brickell and Mathew
1976, p. 6).
Recent studies on the ecological context of reproductive system variation in D. laureola showed that hermaphroditic Iberian
populations were associated with higher ambient temperatures
and lower annual precipitation than gynodioecious ones, and suggested that the abiotic features conditioning gender polymorphism
would be similar at both Northern and Southern regions (Alonso
et al. 2007). Different selective and stochastic processes might
prevail at different geographic locations (e.g., Nilsson and Ågren
2006) and phylogenetic analysis can be essential to unravel, for
instance, if similarity in population reproductive system or average plant size could also reflect genetic similarity between
populations or not. Phylogenetic analysis would be also useful
to identify cryptic species, as in the Wurmbea genus, where the
study of the evolutionary transitions of reproductive system with
a phylogenetic perspective contradicted the monophyly of two
sexually polymorphic taxa, Wurmbea dioica and W. biglandulosa
(Case et al. 2008).
Our analysis of intraspecific genetic relationships between
populations of D. laureola differing in reproductive system aimed
to unravel whether hermaphroditic populations from NE and S
Iberia were closely genetically related to each other and were
placed as a basal group in the phylogeny that could be interpreted
as the relict representatives of a formerly wider distribution range.
That hypothesis would be consistent with hermaphroditism being
the ancestral condition shared by all the analyzed Daphne species
and the macroevolutionary pathway suggested for the whole Angiosperms (Barrett 2002). Results did however indicate that NE
and S hermaphroditic populations were located in different clades
within the phylogenies obtained from both nuclear and plastid
markers and thus do not support a single origin of the populations in which female-sterility mutants are not currently present.
Gynodioecy emerged anyway as the most likely ancestral state
common to all the D. laureola populations sampled, being ubiquitous in the NW clade and also present in the other two sequence
groups (Fig. 5).
Different evolutionary transitions in reproductive system
were traced within each D. laureola clade, with two events of
change from gynodioecy to hermaphroditism and one further
reversion from gynodioecy to hermaphroditism (Fig. 5). Gynodioecy emerged as the most likely condition of the ancestor
common to all the populations sampled at NW Iberian Peninsula,
genetically characterized by ribotype D and haplotypes 1 and 2.
Such an ancestor would have likely survived within this interglacial refugium (see above) from which it would have eventually
recolonized all the Northwestern Iberian mountain ranges including the Western Pyrenees. In contrast, hermaphroditism emerged
as the most likely condition of the ancestor common to Slovenian
and NE Iberian populations sampled (Fig. 5), genetically characterized by the haplotype 6, suggesting another interglacial refugia
of the species located in NE Iberian mountains, that would have
likely followed different recolonizing routes to Central Europe
and the Southern Iberian ranges, where gynodioecy would have
succeeded as the commonest reproductive system in the South
(Alonso et al. 2007). Population dynamics and selective particularities of a complex trait such as reproductive system (Pannell
and Dorken 2006; Case et al. 2008) can likely explain such evolutionary liability although it has been only rarely documented at
intraspecific level (Fénart et al. 2006).
In fragmented landscapes, hermaphroditism and selfing
would be positively selected in recently colonized populations
due to reproductive assurance advantages when the number of
immigrants is low (Pannell 1997; Pannell and Dorken 2006).
A recent metapopulation model further suggests that gene migration among demes can be essential for the dynamics of the
system under nucleo-cytoplasmic inheritance (Dufay and Pannell
2010). At the geographic scale studied here, fragmentation of
the landscape with repeated colonization and extinction of local
populations has likely been the commonest scenario for D. laureola, an understory species of shady mountainous forest, since
the last glaciation. Long-distance dispersal events involving very
few seeds dispersed by birds would allow the colonization of
relatively small and isolated mountains and would be characterized by a reduced genetic diversity within population (Bialozyt
et al. 2006). Interestingly, the Southwestern hermaphroditic population CA is characterized by absence of variability in the 73
adult individuals analyzed (C. Alonso, unpubl. data) for the three
allozyme systems that were known to be polymorphic in several
Southeastern populations (Medrano et al. 2005), thus suggesting
that a long-distance colonizing event could be behind the only
hypothesized reversion from gynodioecy to hermaphroditism observed. Ongoing studies at the Southern Iberian Ranges support
a role of among-population gene flow in the dynamics of gynodioecy (Dufay and Pannell 2010), because the frequency of females
and prevalence of gynodioecy were higher in the central region,
with highly connected populations encompassing a broad altitudinal range, than in the marginal and highly isolated Western
Betic Ranges where the hermaphroditic CA population is located
(A. R. Castilla, C. Alonso and C. M. Herrera, unpubl. data). In addition, Dufay and Pannell (2010) simulations suggest that, when
migration among local populations is absent for a few hundred
generations, the combination of genetic drift and selection will
promote allele fixation, frequently leading to hermaphroditism
within populations, and even at the metapopulation scale if female reproductive advantage is not strong. Such extreme isolation of populations is expected to happen in highly heterogeneous
landscapes such as the small and isolated foothills characteristic of the NE Catalonian Coastal Ranges (Garcı́a-Castellanos
et al. 2003), where all D. laureola populations surveyed were
found to be hermaphroditic and individual plants smaller, presumably reducing geitonogamous selfing and the potential for
a female advantage by inbreeding avoidance (Alonso et al.
2007).
Altogether, our study highlights that plant reproductive system is not necessarily a species-specific trait, but can be a variable, labile feature subject to eco–evo coevolutionary feedbacks
(Kokko and Ló pez-Sepulcre 2007) with hermaphroditism and
selfing being positively selected at the low individual frequencies
characteristic of new colonized populations but with male-sterile
mutants being able to spread in more dense (older) populations
if they reach some reproductive advantage through quantitative
seed production or inbreeding avoidance (Medrano et al. 2005,
and references therein) and are able to maintain some gene flow
among them (Dufay and Pannell 2010). Under this scenario,
species can evolve gynodioecy from hermaphroditism as traditional models predict (Barrett 2002), but gynodioecy can also
revert to hermaphroditism, favored by genetic drift and isolation. A combination of selective and stochastic processes will
determine the course and polarity of evolutionary transitions,
the role of stochasticity being expected to increase with population isolation and limitations to gene flow, as commonly
found at the borders of species distribution ranges (Herrera and
Bazaga 2008). Broad phylogeographic analysis of species that
exhibit among-population variation in reproductive system are
bound to contribute in important ways to understand how the
macroevolutionary history of colonization/extinction and diversification inferred from neutral genetic markers, and the microevolutionary population dynamics of reproductive system, converge to explain the evolution of plant polymorphic reproductive
systems.
ACKNOWLEDGMENTS
We are grateful to G. N. Feliner and J. F. Aguilar for enthusiastically
supporting the onset of this project and introducing us to phylogenetic
techniques; P. Bazaga and B. Guzmán, for assistance with technical issues;
V. Fernández, D. Gó mez, and R. Brus for their help in locating populations
at the Baetic Ranges, Pyrenees, and Slovenia, respectively; and M. Burd
and two anonymous referees for their constructive reviews of the work.
The study was funded by the Spanish Ministerio de Educació n y Ciencia
through the research project CGL2006-01355/BOS, Consejo Superior
de Investigaciones Cientı́ficas (PIE200730/001), and the Consejerı́a de
Innovació n, Ciencia y Empresa, Junta de Andalucı́a, through Excellence
Research Project RNM156-2005.
LITERATURE CITED
Alonso, C., P. Mutikainen, and C. M. Herrera. 2007. Ecological context of
breeding system variation: sex, size and pollination in a (predominantly)
gynodioecious shrub. Ann. Bot. 100:1547–1556.
Ashman, T.-L. 2006. The evolution of separate sexes: a focus on the ecological
context. Pp. 204–222 in L. D. Harder and S. C. H. Barrett, eds. Ecology
and evolution of flowers. Oxford Univ. Press, Oxford, UK.
Bailey, M. F., and L. F. Delph. 2007. A field guide to models of sex-ratio
evolution in gynodioecious species. Oikos 116:1609–1617.
Bailey, M. F., L. F. Delph, and C. M. Lively. 2003. Modeling gynodioecy:
novel scenarios for maintaining polymorphism. Am. Nat. 161:762–776.
Barrett, S. C. H. 2002. The evolution of plant sexual diversity. Nat. Rev. Genet.
3:274–284.
Beaumont, A. J., T. J. Edwards, and F. R. Smith. 2006. The first record of
gynodioecy in a species of Gnidia (Thymelaeaceae) from South Africa.
Bot. J. Linn. Soc. 152:219–233.
Bialozyt, R., B. Ziegenhagen, and R. J. Petit. 2006. Contrasting effects of long
distance seed dispersal on genetic diversity during range expansion. J.
Evol. Biol. 19:12–20.
Brickell, C. D., and B. Mathew. 1976. Daphne. The genus in the wild and in
cultivation. The Alpine Garden Society, Surrey.
Caruso, C. M., and A. Case. 2007. Sex ratio variation in gynodioecious Lobelia
siphilitica: effects of population size and geographic location. J. Evol.
Biol. 20:1396–1405.
Case, A., and S. C. H. Barrett. 2004. Floral biology of gender monomorphism and dimorphism in Wurmbea dioica (Colchidaceae) in Western
Australia. Int. J. Plant Sci. 165:289–301.
Case, A., S. W. Graham, T. D. Macfarlane, and S. C. H. Barrett. 2008.
A phylogenetic study of evolutionary transitions in sexual systems
in Australasian Wurmbea (Colchicaceae). Int. J. Plant Sci. 169:141–
156.
Charlesworth, D. 1999. Theories of the evolution of dioecy. Pp. 33–60 in
M. A. Geber, T. E. Dawson, and L. F. Delph, eds. Gender and sexual
dimorphism in flowering plants. Springer-Verlag, Berlin.
Clement, M., D. Posada, and K. Crandall. 2000. TCS: a computer program to
estimate gene genealogies. Mol. Ecol. 9:1657–1660.
Cornman, R. S., and M. L. Arnold. 2007. Phylogeography of Iris missouriensis (Iridaceae) based on nuclear and chloroplast markers. Mol. Ecol.
16:4585–4598.
Delph, L. F., and S. B. Carroll. 2001. Factors affecting relative seed fitness
and female frequency in a gynodioecious species, Silene acaulis. Evol.
Ecol. Res. 3:487–505.
Desfeux, C., S. Maurice, J.-P. Henry, B. Lejeune, and P.-H. Gouyon. 1996.
Evolution of reproductive systems in the genus Silene. Proc. R. Soc.
Lond. B 263:409–414.
Downie, S. R., and D. S. Katz-Downie. 1996. A molecular phylogeny of
Apiaceae subfamily Apioideae: evidence from nuclear ribosomal DNA
internal transcribed spacer sequences. Am. J. Bot. 83:234–251.
Dufay, M., and J. R. Pannell. 2010. The effect of pollen versus seed flow on
the maintenance of nuclear-cytoplasmic gynodioecy. Evolution 64:772–
784.
Dufay, M., P. Touzet, S. Maurice, and J. Cuguen. 2007. Modelling the maintenance of male-fertile cytoplasm in a gynodioecious population. Heredity
99:349–356.
Fénart, S., P. Touzet, J.-F. Arnaud, and J. Cuguen. 2006. Emergence of gynodioecy in wild beet (Beta vulgaris ssp. maritima L.): a genealogical
approach using chloroplastic nueclotide sequences. Proc. R. Soc. Lond.
B 273:1391–1398.
Galicia-Herbada, D. 2006. Origin and diversification of Thymelaea
(Thymelaeaceae): inferences from a phylogenetic study based on ITS
(rDNA) sequences. Plant Syst. Evol. 257:159–187.
Garcı́a-Castellanos, D., J. Vergés, J. Gaspar-Escribano, and S. Cloetingh.
2003. Interplay between tectonics, climate, and fluvial transport during
the Cenozoic evolution of the Ebro Basin (NE Iberia). J. Geophys. Res.
108:2347–2365.
Gene Codes Corporation. 2007. Sequencher Program. Version 4.8. Ann Arbor,
MI, USA.
Gómez, A., and D. H. Lunt. 2007. Refugia within refugia: patterns of phylogeographic concordance in the Iberian Peninsula. Pp. 155–188 in S.
Weiss and N. Ferrand, eds. Phylogeography of Southern European Refugia. Springer, Dordrecht, The Netherlands.
de Heredia, U. L., J. S. Carrion, P. Jimenez, C. Collada, and L. Gil. 2007.
Molecular and palaeoecological evidence for multiple glacial refugia
for evergreen oaks on the Iberian Peninsula. J. Biogeogr. 34:1505–1517.
Herrera, C. M., and Bazaga, P. 2008. Adding a third dimension to the edge
of a species’ range: altitude and genetic structuring in mountainous
landscapes. Heredity 100:275–285.
Huelsenbeck, J. P., and F. Ronquist. 2001. MrBayes, Bayesian Analysis of
Phylogeny. Bioinformatics 17:754–755.
Kaul, M. L. H. 1988. Gene-cytoplasmic male sterility. Pp. 97–192 in M. L.
H. Kaul, ed. Male sterility in plants. Springer-Verlag, Berlin, Germany.
Kikuzawa, K. 1989. Floral biology and evolution of gynodioecism in Daphne
kamtchatica var. jezoensis. Oikos 56:196–202.
Kokko, H., and E. López-Sepulcre. 2007. The ecogenetic link between demography and evolution: can we bridge the gap between theory and
data? Ecol. Lett. 10:773–782.
Maddison, W. P., and D. R. Maddison. 2007. Mesquite: a modular system for
evolutionary analysis. Version 2.0 http://mesquiteproject.org.
Magri, D., G. G. Vendramin, B. Comps, I. Dupanloup, T. Geburek, D. Gomory,
M. Latalowa, T. Litt, L. Paule, J. M. Roure, et al. 2006. A new scenario for
the Quaternary history of European beech populations: palaeobotanical
evidence and genetic consequences. New Phytol. 171:199–221.
McCauley, D. E., and M. F. Bailey. 2009. Recent advances in the study of
gynodioecy: the interface of theory and empiricism. Ann. Bot. 104:611–
620.
McCauley, D. E., and D. R. Taylor. 1997. Local population structure and sex
ratio: evolution in gynodioecious plants. Am. Nat. 150:406–419.
McCauley, D. E., and M. S. Olson. 2003. Associations among cytoplasmic
molecular markers, gender, and components of fitness in Silene vulgaris,
a gynodioecious plant. Mol. Ecol. 12:777–787.
———. 2008. Do recent findings in plant mitochondrial molecular and population genetics have implications for the study of gynodioecy and cytonuclear conflict?. Evolution 62:1013–1025.
Medrano, M., C. Alonso, and C. M. Herrera. 2005. Mating system, sex ratio,
and persistence of females in the gynodioecious shrub Daphne laureola
L. (Thymelaeaceae). Heredity 94:37–43.
Meusel, H., E. Jäger, S. Rauschert, and E. Weinert. 1978. Vergleichende
Chorologie der zentraleuropäischen Flora. Band III, Gustav Fischer,
Jena.
Morrison, D. A. 2009. A framework for phylogenetic sequence alignment.
Plant Syst. Evol. 282:127–149.
Navajas-Pérez, R., R. de la Herran, G. L. González, M. Jamilena, R. Lozano,
C. R. Rejon, M. R. Rejon, and M. A. Garrido-Ramos. 2005. The evolution of reproductive systems and sex-determining mechanisms within
Rumex (Polygonaceae) inferred from nuclear and chloroplastidial sequence data. Mol. Biol. Evol. 22:1929–1939.
Nieto Feliner, G. 1997. Daphne. Pp. 33–42 in S. Castroviejo, C. Aedo, C.
Benedı́, M. Laı́nz, F. Muñoz Garmendia, G. Nieto Feliner, and J. Paiva,
eds. Flora Iberica. Vol. VIII. Real Jardı́n Botánico, CSIC, Madrid, Spain.
Nilsson, E., and J. Ågren. 2006. Population size, female fecundity, and sex ratio
variation in gynodioecious Plantago maritima. J. Evol. Biol. 19:825–
833.
Olalde, M., A. Herrán, S. Espinel, and P. G. Goicoechea. 2002. White oaks
phylogeography in the Iberian Peninsula. Forest Ecol. Manag. 156:89–
102.
Olmstead, R. G., and J. A. Sweere 1994. Combining data in phylogenetic
systematics: an empirical approach using three molecular data sets in
the Solanaceae. Syst. Biol. 43:467–481.
Pannell, J. R. 1997. The maintenance of gynodioecy and androdioecy in a
metapopulation. Evolution 51:10–20.
Pannell, J. R., and M. E. Dorken. 2006. Colonisation as a common denominator in plant metapopulations and range expansions: effects on genetic
diversity and sexual systems. Land. Ecol. 21:837–848.
Pedrol, J. 1997. Thymelaea Mill. Pp. 42–69 in S. Castroviejo, C. Aedo, C.
Benedı́, M. Laı́nz, F. Muñoz Garmendia, G. Nieto Feliner, and J. Paiva
eds. Flora Iberica. Vol. VIII. Real Jardı́n Botánico, CSIC, Madrid, Spain.
Posada, D., and K. Crandall. 1998. Modeltest: testing the model of DNA
substitution. Bioinformatics 14:817–818.
Ramsey, M., G. Vaughton, and R. Peakall. 2006. Inbreeding avoidance and
the evolution of gender dimorphism in Wurmbea biglandulosa (Colchicaceae). Evolution 60:529–537.
Sang, T., D. J. Crawford, and T. F. Stuessy. 1997. Chloroplast DNA phylogeny,
reticulate evolution and biogeography of Paeonia (Paeoniaceae). Am. J.
Bot. 84:1120–1136.
Schultz, S. T. 1994. Nucleo-cytoplasmic male sterility and alternative routes
to dioecy. Evolution 48:1933–1945.
Storchova, H., and M. S. Olson. 2007. The architecture of the chloroplast psbA-trnH non-coding region in angiosperms. Plant. Syst. Evol.
268:235–256.
Swofford, D. L. 2002. PAUP: phylogenetic analysis using parsimony and other
methods. Version 4. Sinauer Associates, Sunderland, MA, USA.
Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins.
1997. The ClustalX windows interface: flexible strategies for multiple
sequence alignment aided by quality analysis tools. Nucleic Acids Res.
25:4876–4882.
Touzet, P., and L. F. Delph. 2009. The effect of breeding system on polymorphism in mitochondrial genes of Silene. Genetics 181:631–644.
Vamosi, J. C., S. P. Otto, and S. C. H. Barrett. 2003. Phylogenetic analysis
of the ecological correlates of dioecy in angiosperms. J. Evol. Biol.
16:1006–1018.
Van Der Bank, M., M. F. Fay, and M. W. Chase. 2002. Molecular phylogenetics
of Thymelaeaceae with particular reference to African and Australian
genera. Taxon 51:329–339.
Weiblen, G. D., R. K. Oyama, and M. J. Donoghue. 2000. Phylogenetic
analysis of dioecy in monocotyledons. Am. Nat. 155:46–58.
Weller, S. G., W. L. Wagner, and A. K. Sakai. 1995. A phylogenetic analysis
of Schiedea and Alsinidendron (Caryophyllaceae: Alsinoideae): implications for the evolution of breeding systems. Syst. Bot. 20:315–337.
Weller, S. G., and A. K. Sakai. 1999. Using phylogenetic approaches for
the analysis of plant breeding system evolution. Annu. Rev. Ecol. Syst.
30:167–199.
White, T. J., T. Bruns, S. Lee, and J. Taylor. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenies. Pp. 315–322
in M. Innis, D. Gelfand, J. Sninsky, and T. White, eds. PCR protocols: a
guide to methods and applications. Academic Press, San Diego, CA.
Associate Editor: M. Burd
Supporting Information
The following supporting information is available for this article:
Appendix S1. Alignment of all the ITS sequences used in this study, including their GenBank accession numbers.
Appendix S2. Alignment of all the ndhF sequences used in this study, including their GenBank accession numbers.
Appendix S3. Alignment of all the psbA-trnH intergenic spacer sequences used in this study, including their GenBank accession
numbers. Note that the 5 bp inversion at position 38 was coded as a single change.
Supporting Information may be found in the online version of this article.
Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting information supplied by the
authors. Any queries (other than missing material) should be directed to the corresponding author for the article.